Andrew Howard | Business Strategist & Coach | NZ + UK

Executive Summary: Traditional risk models fail in today’s fragmented world. Businesses investing billions in geopolitical intelligence partnerships (GardaWorld’s $14B recapitalisation, Crisis24 + Palantir) are not buying predictions. They’re buying a faster reaction time. AI helps spot patterns, but adaptability beats forecasting. Build redundancy, learn from near-misses, and accept uncertainty as permanent.

Core insights:

• Geopolitical risk consulting shifted from niche to necessity as global systems fragment

• AI-driven intelligence (Crisis24-Palantir partnership) democratises government-grade analysis for corporates

• Prediction paradox: modelling risk changes the system itself (Strait of Hormuz scenario shows self-inflicted harm)

• Winners build rapid adaptation capacity, not better forecasts

• Corporate responsibility now includes monitoring distant geopolitical events with local impact

Why Traditional Risk Models Fail Now

I’ve spent three decades watching businesses try to predict the future. Most get it wrong.

Not from carelessness. Not from lack of data. They fail because the world has become fundamentally harder to read.

Old playbooks assumed logic. Markets moved in patterns. Geopolitics stayed background noise. Supply chains bent but held. You modelled risk with reasonable confidence.

That world ended.

Key Point: Risk models built on historical patterns break when the variables themselves change.

What’s Driving the Shift to Geopolitical Intelligence

I’m watching something shift in how companies approach uncertainty. Started quietly. Accelerating fast.

Geopolitical risk consulting moved from niche to necessity. The numbers prove it. GardaWorld, one of the largest security firms globally, secured $14 billion in recapitalisation for 2025. This is transformation money, not maintenance capital.

Their intelligence division, Crisis24, partnered with Palantir Technologies. Palantir built data platforms for intelligence agencies. Now they’re turning those capabilities towards corporate clients who need to understand global disruption before operations get hit.

This is not insurance. This is foresight.

Key Point: The Crisis24-Palantir partnership signals the democratisation of government-grade intelligence for corporate risk management.

How Globalisation Fragmentation Creates New Vulnerabilities

I’ve sat in boardrooms where executives presented confident risk matrices. Neat columns. Probability percentages. Impact scores. Everything colour-coded.

Then something happens outside the grid.

Traditional risk assessment assumes the future roughly resembles the past. Maps known variables. We’re living through a period where variables themselves change.

Globalisation is fragmenting. Supply chains, once permanent, became political bargaining chips. Energy flows powering entire economies get weaponised overnight. Technology connecting us creates new vulnerabilities.

Businesses today face questions predecessors never asked:

• What happens when our primary shipping route becomes a conflict zone?

• How do we operate when sanctions reshape our supplier network?

• What’s our contingency when critical infrastructure becomes a target?

• How do we protect people in markets where stability evaporates?

These are not theoretical exercises.

Key Point: Fragmented globalisation means supply chains, energy flows, and technology infrastructure become geopolitical weapons, creating risks that traditional models never anticipated.

The Role of AI in Risk Assessment

Artificial intelligence entered the conversation as a potential solution. The pitch compels: feed the machine enough data, and you get granular risk assessments humans miss.

I’m cautiously optimistic about AI here. The Crisis24-Palantir partnership suggests we’re moving towards democratised intelligence. Capabilities once exclusive to government agencies become available to corporates.

Powerful. But comes with a warning.

The Prediction Paradox

Predicting risk is itself risky.

The Global Risk Forecast 2026 includes a scenario where sustained efforts to block the Strait of Hormuz destabilise energy markets. Shipping costs spike. Insurance premiums explode. Supply chains reroute.

The twist: the action inflicts massive damage on Iran’s own economy. The country creating instability harms itself more than its targets.

This is the paradox. When you model risk, you’re not observing a system. You’re participating in it. Predictions become self-fulfilling. Or they trigger countermeasures, changing the outcome entirely.

AI processes vast data. Spots patterns humans miss. But AI cannot account for the irrational, the emotional, or unintended consequences of its own predictions being acted upon.

Key Point: AI enhances pattern recognition but cannot predict how rational actors respond to forecasts, creating a feedback loop where predictions alter outcomes.

What Works: Building Adaptive Capacity

I’ve seen businesses survive impossible situations. Seen well-resourced companies collapse when conditions shift.

The difference was not better forecasting. Faster adaptation.

Four Characteristics of Resilient Organisations

1. Build redundancy into critical systems

Not because you predict specific failures. Because you assume something will fail.

2. Maintain optionality

Multiple suppliers. Flexible contracts. Diverse markets. Pay the premium for avoiding lock-in to a single path.

3. Learn from near-misses

Most organisations study disasters. Smart ones study close calls and ask what they revealed about vulnerabilities.

4. Accept uncertainty as permanent

Watch how many strategic plans assume a return to normal. There is no normal coming. There’s the next phase of instability.

Key Point: Resilient businesses build redundancy, maintain optionality, learn from near-misses, and treat uncertainty as permanent rather than waiting for stability to return.

Why Corporate Responsibility Now Includes Geopolitics

The definition of what businesses need to monitor is expanding rapidly.

Ten years ago, a manufacturing company worried about commodity prices and labour costs. Today, the same company needs to understand Middle Eastern geopolitics, Chinese industrial policy, European energy security, and American trade strategy.

Global interconnectedness means local actions have distant consequences. A port strike in one country cascades through supply chains across three continents. A policy decision in one capital reshapes investment flows globally.

This creates a challenge for business owners already stretched thin. You’re running operations, managing people, and hitting financial targets. Now you’re also a geopolitical analyst?

The honest answer: yes. Or you need access to the analysis.

This is why I’m watching firms like Crisis24 with interest. They’re not selling security services in the traditional sense. They’re selling context. Helping businesses understand the operating environment beyond immediate markets.

Key Point: Global interconnectedness forces businesses to monitor geopolitical events previously considered distant, because local actions now cascade across continents.

The Data Bias Problem in AI Risk Assessment

As AI becomes central to risk assessment, we need to address something uncomfortable: these systems are only as good as their data sources and algorithmic assumptions.

I worry about businesses outsourcing judgment to black-box systems they don’t fully understand. AI identifies correlations but cannot always explain causation. Flags anomalies, but cannot tell you which ones matter.

There’s also the question of whose perspective shapes the model. When your risk assessment AI is trained primarily on Western data sources, how well does it understand risk in Asian or African markets? When optimising for shareholder value, does it adequately weigh risks to employees or communities?

These are not philosophical questions. They have practical implications for how you allocate resources and make decisions.

Key Point: AI risk models reflect their training data biases, raising questions about geographic coverage and stakeholder weighting in algorithmic decision-making.

A Practical Framework for Navigating Uncertainty

Here’s a framework worth considering for handling this level of uncertainty:

1. Start with what you control

You cannot predict the next geopolitical crisis, but you can stress-test your supply chain. You cannot prevent market disruption, but you can build cash reserves. You cannot eliminate risk, but you can reduce exposure to single points of failure.

2. Invest in intelligence, not information

Data is everywhere. Intelligence is data filtered, analysed, and contextualised for your specific situation. Partnerships with specialised firms add value here. They watch signals you don’t have time to monitor.

3. Build organisational muscle for rapid response

When something breaks, how fast do you pivot? Do you have decision-making processes working under pressure? Have you practised scenario planning with your leadership team? Build this capability before you need it.

4. Learn systematically from every disruption

After-action reviews should not only happen after disasters. Every supply delay, market shift, and unexpected event is data. What did it reveal about your assumptions? What worked in your response? What failed?

5. Accept that perfect information is not coming

You’re going to make decisions with incomplete data. This is not a planning failure. This is the operating environment. Get comfortable with it.

Key Point: Focus on controllable factors, invest in contextualised intelligence, build rapid response capacity, learn from all disruptions, and accept imperfect information as standard.

What the Industry Won’t Admit

Here’s something most consultants will not say: nobody knows what’s coming.

The experts don’t know. The AI doesn’t know. The intelligence agencies don’t know. We’re all working with probabilities and educated guesses.

What separates successful businesses from failing ones is not certainty. Resilience. The ability to absorb shocks, adapt quickly, and keep operating when conditions change.

Companies investing heavily in geopolitical risk consulting are not doing it because they think they can predict the future. They’re doing it to shorten reaction time when the future arrives unexpectedly.

Key Point: Investment in geopolitical intelligence is about reducing reaction time, not achieving perfect foresight.

What to Expect Next

I expect the trend towards sophisticated risk assessment to accelerate. More partnerships between security firms and technology companies. More AI-driven analysis. More corporate investment in intelligence capabilities.

This creates advantages for larger organisations with resources to invest. Also creates opportunities for smaller businesses to access these tools through partnerships or platforms.

But technology alone will not solve the problem. The human element remains critical. Judgement. Experience. The ability to read between the lines and sense when something feels wrong.

The businesses thriving will not be the ones with the best predictions. They’ll be the ones moving fastest when predictions prove wrong.

Because in a chaotic world, speed beats accuracy. Adaptation beats forecasting. Resilience beats optimisation.

The map doesn’t work. Businesses accepting this and learning to navigate without one will still be operating when the next crisis hits.

There will be a next crisis. There always is.

Frequently Asked Questions

What is geopolitical risk consulting?

Geopolitical risk consulting provides businesses with an analysis of how political events, conflicts, and policy changes in other countries affect their operations. Firms like Crisis24 monitor global disruptions and help companies understand threats to supply chains, personnel safety, and market access before these issues impact operations.

Why is traditional risk modelling failing businesses now?

Traditional risk models assume future conditions will resemble past patterns. Today’s environment features fragmenting globalisation, weaponised supply chains, and rapidly changing variables. When the underlying structure of global systems shifts, historical data becomes a poor predictor of future risk.

How does the Crisis24-Palantir partnership change corporate risk management?

This partnership brings intelligence-agency-grade data analysis capabilities to corporate clients. Palantir’s platforms, originally built for government intelligence work, now process global disruption data for businesses. This democratises access to sophisticated analysis previously available only to governments.

What is the prediction paradox in risk assessment?

The prediction paradox occurs when forecasting risk changes the system being predicted. When businesses act on risk forecasts, their actions alter outcomes. Predictions become self-fulfilling or trigger countermeasures. The Strait of Hormuz scenario illustrates this: blocking the strait would harm Iran’s economy more than the target countries, showing how destabilising actions backfire.

What should small businesses do about geopolitical risk when they lack resources for intelligence partnerships?

Start with controllable factors. Stress-test supply chains. Build cash reserves. Reduce single points of failure. As platforms emerge, smaller businesses will access intelligence tools through partnerships or subscription models. Focus on building adaptive capacity and learning systematically from disruptions.

How does AI improve risk assessment?

AI processes vast amounts of data and identifies patterns humans miss. AI-driven systems analyse multiple variables simultaneously and flag anomalies. The limitation: AI cannot account for irrational behaviour, emotional responses, or unintended consequences of predictions being acted upon.

What is the biggest shift in corporate responsibility regarding risk?

Businesses now need to monitor geopolitical events in distant regions because global interconnectedness means local actions cascade across continents. A port strike, energy policy, or conflict in one region reshapes supply chains and investment flows globally. Corporate responsibility expanded to include understanding distant political dynamics with local operational impact.

How to build organisational adaptability

Build redundancy in critical systems. Maintain multiple suppliers and flexible contracts. Practise scenario planning with leadership teams. Develop decision-making processes that function under pressure. Learn systematically from near-misses, not only disasters. Treat uncertainty as permanent rather than temporary.

Key Takeaways

• Traditional risk models break when global systems fragment because they rely on historical patterns during a period of structural change.

• The $14 billion GardaWorld recapitalisation and Crisis24-Palantir partnership signals geopolitical intelligence moving from niche to corporate necessity.

• AI enhances pattern recognition in risk assessment but cannot predict how its own forecasts alter actor behaviour, creating a prediction paradox.

• Resilient businesses prioritise rapid adaptation over accurate forecasting by building redundancy, maintaining optionality, and learning from near-misses.

• Global interconnectedness expanded corporate responsibility to include monitoring distant geopolitical events because local actions cascade across supply chains and markets.

• Investment in geopolitical intelligence aims to reduce reaction time when disruptions occur, not achieve perfect foresight.

• Speed, adaptation, and resilience outperform accuracy and optimisation in chaotic operating environments.